Rössler  (1961:170) compares to it Ugr 
γnb - ‘grapes’(?), considering the Akk term a WS borrowing - but ḫanābu is reliably attested in OB, which makes the hypothesis rather problematic. However, Akk. ḫ is more likely to correspond to *γ than to *ˁ. 
See also PS 
*ˁinab-